Category Archives: Tech

Tech

Optical Mapping and Optical Coherence Tomography

In my last few posts, I’ve mentioned optical mapping and optical coherence tomography (OCT). Given just the names it’s not actually obvious that they’re entirely different things. Thus, I thought I should explain them.

Optical Mapping

“Optical Mapping” is short for something like “Optical mapping of voltage on the surface of the heart using fluorescence from potentiometric dyes”. The short version is ambiguous, as it could mean optical mapping of all kinds of stuff, and could in fact be synonymous with OCT. When cardiac electrophysiologists refer to an optical mapping study, what that usually involves is the following.

A heart is mounted in a clear chamber of some kind, and instrumented appropriately for whatever experiment is being done (say, defibrillation). A dye is then injected into the fluid flowing through the heart, typically “Tyrode’s” solution, which is kind of like blood without proteins and blood cells. This dye is fluorescent, meaning that when a certain color of light shines on it, it absorbs that light but then releases light of a different color. The cool (and useful) thing is that the amount of light released depends on the voltage gradient across the dye. It is designed to sit across the membrane of a cell, so by measuring changes in light released, it’s possible to measure changes in the electrical potential across the cell membrane, and to thereby monitor excitation, arrhythmia, and so on without electrodes. Using a fast, high-resolution camera, it is thereby possible to measure electrical activity in the heart from many more points, and at much better resolution, than using a bunch of electrodes. The major downside of optical mapping is that it is essentially limited to surfaces. Electrodes are still needed to probe the depths of the myocardium. Alternatively, optical mapping experiments can be combined with simulation experiments. We contend that if the sufrace activity of the simulation matches the suface activity of the experimental preparation “reasonably” well, the simulated activity within the walls should be a good approximation of what’s going on inside the experimental heart.

Optical Coherence Tomography

Typically referred to as OCT, the easiest way to explain it is that it’s like ultrasound imaging, but with light. In reality it gets a little more complicated. Wikipedia has a pretty good explanation. We’re using it to scan the thin right-ventricular free wall of the rabbit heart, and are then going to make and use computational models based on the resulting data. To do that, we have to segment the images (determine what is tissue, what is not), and then analyze them to create a finite element mesh.

Follow-up on PLoS ONE Ratings, from PLoS

I mentioned yesterday the new rating system/software launched by PLoS one. This was precipitated by an announcement on the software/technical side. Today they have an article on the PLoS blog about the new rating system, encouraging people to rate articles.

Here’s the intro from the post:

I’ve been waiting to write this Blog posting for a while and now I can. As from today PLoS ONE has a user rating system for its articles. All users can now rate articles in three subjective categories: Insight, Reliability and Style. We have made the tool, now we need you to come and use it.

User rating is a very common feature of websites these days, be it for movies, books, blog posts, pretty much anything. What user rating allows is a quick and easy survey of a communities opinion. Despite the obvious advantages to hard pressed scientists trying to get to grips with a vast literature this simple system hasn’t been much applied to scientific papers up to this point.

The major exception to this is probably Faculty of 1000, which has been providing ratings for papers for many years, but that is not based on the opinion of a whole community but only the thoughts of a select few.

So what will this new rating system look like? Well, if you go to any of the six hundred or so papers that PLoS ONE has so far published and look in the right had column you will see a little box containing five small stars. Those indicate the overall aggregate rating of the paper based on individual ‘votes’ from individual users.

What’s interesting about this is that it’s a little different from a citation index, the main way that articles are scored. You see, normally scientific articles are given a ranking or score based on how much they are cited in other articles. This is a pretty good idea, but it neglects sort of “terminal” articles — that is, articles that mark the end of most investigation into a particular niche. These articles may nonetheless be extremely interesting or useful, but never garner a large citation index. Furthermore, articles of interest to people in other fields, or even the general public, will never garner any indication of said interest or popularity under the conventional system. The occasional exception might be popular science articles inspired by new publications in Nature or Science.

With the emergence of an article rating system, that may change. People can read and rate articles without having to write an entire manuscript. People can leave comments on articles without drafting (and having accepted) an “official” editorial or response in a major academic journal. Things are getting a lot more interesting, and quickly.

As the title of the article suggests, Rate Early, Rate Often.

More on open science: PLoS ONE

I wrote last month about my desire for more open access in science, and how PLoS was leading the way. I also said I had some ideas about what more could be done.

It looks like PLoS ONE stole them.

Of course, I jest.

Obviously they’ve been working on this for a while, and I missed it. I only know about it because I’m now subscribed to their RSS feed. This is the framework upon which the future of science publishing will be built. There are many PLoS (Public Library of Science, by the way) journals — PLoS ONE seems to aim to be the open-access equivalent of Science or Nature.

Anyone is allowed to register and comment on or rate articles (with some caveats). They even have guidelines for rating articles. The initial review they mention by an editor and perhaps a few reviewers keeps out the quacks, but anyone is allowed to point out weaknesses in the article or make suggestions on it. It reverses my original idea a bit, in which the article would not be “published” until it had passed a vote by public reviewers, but it is perhaps a more functional model.

I am planning to register and see about reviewing and commenting on some articles. Are you?

Ease of Install: Vista 0, FC6 1

Our new post-doc requested both Vista and Fedora Core on his workstation. I started off with Vista. First, it didn’t see any disks. Then, I provided it with drivers for the RAID controller, and it saw the two volumes. However, it refused to install on either, saying they “weren’t suitable” for Vista. Finally, I had to un-hook the disks from the hardware RAID controller, and hook them straight up to the system board.

Contrast this with Fedora Core 6 installation: it finds both volumes, and installs onto either just fine.

Yet, FC6 is free, Linux is supposedly more difficult to install, and Vista Ultimate costs an arm and a leg.

I am annoyed.

ADDENDUM: The Vista install just hung, hard. Good job guys.

Google Reader – Twice Rejected

Back when Google Reader first came out, I decided to check it out. After all, Google’s pretty good at online stuff. Maybe they’d nail the online RSS aggregator the first time around.

Nope.

However, I did a presentation last Friday for the BME grad students here on using PubMed with RSS, and used Google Reader as an example online reader. In the process, I ended up trying it out again. It was looking much better than last time. The interface for reading feeds on my Treo is great as well. However, it still has one fatal flaw.

Yesterday, Amanda asked me about a LiveJournal post she had made. Amanda’s LiveJournal, like mine, is now “friends-only” and requires authentication to access the feed. I hadn’t seen her post. In fact, I hadn’t seen any LJ posts from friends-only journals all weekend. It turns out that Google Reader still doesn’t support authenticated feeds. I tried faking it out with TinyURL and a Yahoo! pipe thing that someone made, to no avail. Back to Gregarius I go.

It’s really too bad. I loved the updated Google Reader interface. However, until they allow authenticated feeds, I can’t migrate.